This repository contains the knowledge-based resources supporting the BeAware Framework for adaptive behavior trees in robotic systems. The repository is organized into two main folders that serve different purposes:
Purpose: Reference implementation and paper examples
Contains the canonical knowledge files that support the research paper "Knowledge-based Execution Configuration for Adaptive Behavior Trees." This folder includes:
- Behavior Trees (BTs): Base templates defining the structure and flow for the example scenario's adaptive behaviors
- Ontologies:
beaware_ontology.owl- The foundational BeAware Framework ontologyapplication_ontology_example.owl- An instantiated ontology for the paper's example scenario
- PDDL: Domain files formalizing the planning aspects of the example scenario
- Videos: Demonstration videos illustrating the execution of behaviors with and without execution configuration flavors
This folder provides the minimal, clean examples discussed in the paper for understanding the core concepts and methodology.
Purpose: Full implementation and development resources
Contains the complete, production-ready knowledge base for a kitchen manipulation scenario with the TIAGo robot. This folder includes:
- Behavior Trees (BTs): Comprehensive behavior tree libraries for real-world robot execution
- Ontologies: Full semantic knowledge structures with robot configurations, environment models, and object definitions
- PDDL: Complete planning domains for task-level planning with PlanSys2 integration
- Configuration: YAML files for the Knowledge Manager runtime configuration
- Outputs: Generated and runtime files (expanded behaviors, reasoned ontologies, execution traces)
- ROS Outputs: Integration outputs from ROS-based robot execution
This folder provides the full implementation details needed to deploy the framework in real robotic applications, including the specific a use case with a TIAGo robot.